Cloudera Docs

Internal Network File System on OCP

Learn about backing up and uninstalling an internal NFS server on OpenShift Container Platform.

Backing up Project Files and Folders

The block device backing the NFS server data must be backed up to protect the Cloudera Machine Learning project files and folders. The backup mechanism would vary depending on the underlying block storage system and backup policies in place.

  1. identify the underlying block storage to backup, first determine the NFS PV:

    $ echo `kubectl get pvc -n cml-nfs -o jsonpath='{.items[0].spec.volumeName}'`
       pvc-bec1de27-753d-11ea-a287-4cd98f578292

  2. For Ceph, the RBD volume/image name is the name of the dynamically created persistent volume (pvc-3d3316b6-6cc7-11ea-828e-1418774847a1).

Ensure this volume is backed up using an appropriate backup policy. The Backup/Restore feature is not yet supported for workspaces with an internal NFS.

Uninstalling the NFS server on OpenShift

Uninstall the NFS server provisioner using either of the following commands.

Use this command if the NFS server provisioner was installed using oc and yaml files: 
$ oc delete scc nfs-scc
$ oc delete clusterrole cml-nfs-nfs-server-provisioner 
$ oc delete clusterrolebinding cml-nfs-nfs-server-provisioner 
$ oc delete namespace cml-nfs
Use this command if the NFS server provisioner was installed using Helm: 
$ helm tiller run cml-nfs -- helm delete cml-nfs --purge
$ oc delete scc nfs-scc securitycontextconstraints.security.openshift.io "nfs-scc" deleted

Storage provisioner change

In CDP 1.5.0 on OCP, Cloudera Machine Learning has changed the underlying storage provisioner for internal workspaces.

Any new internal workspace on 1.5.0 will use CephFS as the underlying storage provisioner. A storage class named "ocs-storagecluster-cephfs" with csi driver set to "openshift-storage.cephfs.csi.ceph.com" must exist in the cluster for new internal workspaces to get provisioned. Each workspace will have separate 1 TB internal storage.

Internal workspaces running on PVC 1.4.0/1 use NFS server provisioner as the storage provisioner. These workspaces when upgraded to 1.5.0 will continue to run with NFS Provisioner. However, NFS server provisioner is deprecated now and will not be supported in 1.5.1 release. So, customers are expected to migrate their 1.5.0 upgraded workspace from NFS server provisioner to CephFS if they want to continue using the same workspace in 1.5.1 as well. If not, then customer should create a new 1.5.0 workspace and migrate their existing workloads to that before 1.5.1 release.

Portworx storage provisioner

Cloudera Machine Learning now supports Portworx as storage backend for workspaces. New Cloudera Machine Learning Workspaces can be configured to use Portworx volumes for storing data.

Steps:

  1. Create a Portworx storage class in your OCP cluster manually. This storage class should have the provisioner field set to either pxd.portworx.com or kubernetes.io/portworx-volume.

    Sample storage classes are:

    apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: sample-portworx-storage-class-k8s
parameters:
repl: "3"
allow_all_ips: "true"
provisioner: kubernetes.io/portworx-volume
reclaimPolicy: Delete
allowVolumeExpansion: true
volumeBindingMode: Immediate 
      
---
      
      
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: sample-portworx-storage-class-pxd 
parameters:
repl: "2"
sharedv4: "true"
sharedv4_svc_type: ClusterIP
allow_all_ips: "true"
provisioner: pxd.portworx.com
reclaimPolicy: Retain
allowVolumeExpansion: true     
volumeBindingMode: Immediate
      
---
      
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
name: sample-portworx-storage-class-pxd 
parameters:
repl: "2"
sharedv4: "true"
sharedv4_svc_type: ClusterIP
allow_ips: 10.17.0.0/16
provisioner: pxd.portworx.com
reclaimPolicy: Retain
allowVolumeExpansion: true     
volumeBindingMode: Immediate
  2. In the Cloudera Machine Learning Workspace provisioning page, enable Set Custom NFS Storage Class Name". Enter the name of storage class created manually above into the Storage Class Name text box.
  3. Enter the rest of the workspace details and submit. This will create a Portworx-backed Cloudera Machine Learning Workspace.